Cell-Free Massive MIMO System for Indoor Industrial IoT Networks

This paper investigates the application of Cell-Free massive MIMO (CF-mMIMO) in indoor industrial environments, a key technology for ensuring reliability and spectral efficiency (SE) in 5G and beyond. We implement two schemes, centralized and distributed, to accommodate different levels of cooperation between access points (APs) and central processing units. We introduce an effective AP selection method and a pilot assignment scheme to mitigate pilot contamination (PC). We examine the factors influencing uplink SE, such as vertical distance and PC. Moreover, we employ a scalable CF-mMIMO model to reduce the complexity without compromising SE. Numerical results demonstrate the superior performance of the centralized CF-mMIMO, showcasing improved SE in uplink and downlink. The analysis shows that smaller vertical distances enhance SE, even in the presence of PC. Furthermore, the results demonstrate the effectiveness of the scalable model, revealing that connecting the user to fewer APs does not significantly degrade performance while reducing complexity based on a chosen threshold. The threshold value is the maximum allowable difference between the large-scale fading coefficients of a candidate AP and the main AP for each user to serve as an AP. Selecting a threshold value of \Psi = -40 dB reduces the maximum number of serving APs to 30% of the total, significantly lowering complexity while decreasing the SE by only 1 bits/s/Hz at most. The results reveal that serving UE with only one AP does not significantly compromise performance for a small number of UEs. However, it affects performance when the number of UEs is large.